This is the mail archive of the
gcc@gcc.gnu.org
mailing list for the GCC project.
Re: reload displacement optimization
- To: richard dot earnshaw at arm dot com
- Subject: Re: reload displacement optimization
- From: Jeffrey A Law <law at cygnus dot com>
- Date: Tue, 10 Mar 1998 11:08:24 -0700
- cc: Richard Henderson <rth at cygnus dot com>, egcs at cygnus dot com, rearnsha at arm dot com
- Reply-To: law at cygnus dot com
In message <199803101259.MAA21273@sun52.NIS.cambridge>you write:
> Can't we have a RELOAD_LEGITIMIZE_ADDRESS() in the same way that we have LE
> GITIMIZE_ADDRESS(). This could then be a tuning option which will be used
> if it produces valid insns, but thrown out otherwise.
Joern, myself and now rth have done basically the same thing over
the years.
The ability to specify a machine dependent way to reload addresses
is badly needed.
My biggest concern with Joern's version was documenting exactly
reload expected the macro to do in terms of calling push_reload,
rtx sharing and such. Revisiting Joern's patch has been on my
todo list for a while.
rth -- maybe you should look at Joern's patch work forward from
it.
jeff